Factors regulating serial dependencies in the discharge patterns of nerve cells: a computer simulation study in relation to dorsal spinocerebellar tract neurones.

This study was undertaken to investigate the factors which may contribute to the strong negative correlation between successive interspike intervals which characteristically occur in the discharge of dorsal spinocerebellar tract (DSCT) neurones. This correlation is reflected by a first-order serial correlation coefficient (R1) which is often less than --0.6(17). This strongly negative correlation would be expected to increase the rate at which the DSCT neurone transmits information from peripheral receptors. A model of a DSCT neurone has been developed and simulated on a digital computer. It was found that the discharge pattern of the model could be very similar to that of a real DSCT neurone and the value of R1 could be less than --0.6. In order for such values of R1 to be generated, a number of conditions had to occur simultaneously. These were that the discharge of the excitatory presynaptic fibres impinging on the neurone had to be fairly regular, that the overall range of firing frequencies on these fibres had to be rather small and that the mean firing rate of the neurone had to be about double that of each presynaptic fibre. Provided these 3 conditions were met, the precise properties of the model were relatively unimportant. Parameter values derived from biophysical studies of real DSCT neurones fulfilled these conditions.